Concepts of various machine learning methods

Reinforcement Learning ):

The so-called reinforcement learning is the learning of intelligent systems from environment to behavior ing to enable the reward signal (strongSignal conversion)FunctionValueMaximum, reinforcement learning is different from supervised learning in connection learning, mainly manifested in the signal of teachers, in reinforcement learning, enhanced signals provided by the Environment evaluate the quality of the generated actions (usually scalar signals), rather than telling the Reinforcement Learning System) how to generate the correct action. Due to the small amount of information provided by the external environment, the company must learn from its own experiences. In this way, recursive will gain knowledge in the action-evaluation environment and improve the Action Plan to adapt to the environment.

Incremental Learning ):

Learn new knowledge and never forget old knowledge.

1)New knowledge can be learned from new data;

2) Previously processed data does not need to be processed repeatedly; 

3) Only one Training Observation sample is seen and learned each time;

4) While learning new knowledge, you can save most of the previously learned knowledge;

5) -After the learning is completed, the training and observation samples are discarded;

6) The learning system does not have prior knowledge about the entire training sample.

Online learning ):

Transfer learning ):

Migration learning can migrate knowledge from existing data to help you learn in the future. Transfer Learning (transfer) Learning) aims to use the knowledge learned from an environment to help learning tasks in the new environment. Therefore, migration learning does not make the same Distribution Assumption as traditional machine learning.

Batch learning ):

Concepts of various machine learning methods